Re: [Swsusp-devel] Re: nvidia's driver and swsusp (need help w/ n forc e2 mobo)

Re: [Swsusp-devel] Re: nvidia's driver and swsusp (need help w/ n forc e2 mobo)

[Karol Kozimor]

Thus wrote Terence Ripperda:
> I tried hibernating/resuming while in X. but I don't see any acpi
> calls coming through to our driver via the pci driver model. is this
> expected?

I'm no PM expert, but I'm getting the idea that you're wrong at your
principles. 

To begin with, there's no such thing as acpi calls. There's driver model in
2.6 (i.e. pci_module_init() / pci_register_driver()), and there's 2.4's
lack of thereof (i.e. pm_register()). As far as I know, those two
interfaces have nothing specifically to do with ACPI nor APM. 

I think the proper way to conditionalize this in the code is to check for
CONFIG_PM as the prerequisite and KERNEL_2_6/!KERNEL_2_6 to choose between
those interfaces (and only that: bear in mind swsusp2 and pmdisk require
neither APM nor ACPI). Perhaps you're building with only CONFIG_PM defined?

Hopefully others will be able to clarify this.
Best regards,

-- 
Karol 'sziwan' Kozimor
sziwan@hell.org.pl

Re: [Swsusp-devel] Re: nvidia's driver and swsusp (need help w/ n forc e2 mobo)

[Pavel Machek]

Hi!

> > I tried hibernating/resuming while in X. but I don't see any acpi
> > calls coming through to our driver via the pci driver model. is this
> > expected?

If you have suspend/resume methods in struct pci_device, those should
be called. If they are not, something is very wrong... Take a look on
for example drivers/net/b44.c -- that implements suspend/resume.
								Pavel
-- 
People were complaining that M$ turns users into beta-testers...
...jr ghea gurz vagb qrirybcref, naq gurl frrz gb yvxr vg gung jnl!